7
parties’ expenditures on the district vote share is to use the expenditures of the main rival
challenger. For the majority of LDP candidates in the 2000 Lower House election, the
main rivals were usually incumbents from the DPJ or another opposition party. Likewise,
DPJ candidates faced the most competition from incumbents and newcomers from the
LDP.
To examine the effects of own and rival expenditures upon SMD vote shares, I
will empirically test the following two hypotheses using campaign expenditures for all
LDP and DPJ candidates competing in the 2000 Lower House election:
Hypothesis 1: Politicians’ SMD vote shares are positively correlated with their
own expenditures.
Hypothesis 2: Politicians’ SMD vote shares are negatively correlated with the
expenditures of their main interparty rival .
As the majority of candidates in the LDP and DPJ are dual-listed in the SMD and PR
tiers, the relationship between expenditures and SMD vote shares are likely to respond in
a similar fashion to both hypotheses.
6
However, it is necessary to control for other
candidate and district-specific characteristics that may influence the expected empirical
relationships.
4. Candidate Characteristics
To examine the empirical relationship between expenditures and votes in the new
system, it is necessary to control for candidate-specific characteristics. In Japan as well as
6
Specifically, the consequence of dual-listing and the use of the best-loser provision for candidates ranked
at the same party list position should be to increase the importance of the SMD contest for both of these
parties.
Convention